Orientation correction for chest images

Abstract

This report presents an automatic procedure that determines the orientation of computed radiography (CR) chest images and rotates them to a standard position to be viewed by radiologists. As an input, CR images of a normalized size of 1,000 x 1,000 or 2,000 x 2,000 pixels are used. The analysis is performed in three steps. First, the orientation of the spine within the image is determined. Then, a function searches for upper extremities and the subdiaphragm. Finally, the lungs are extracted and their areas are compared. This indicates whether the image needs to be y-axis flipped. These three steps set the value of three parameters on the basis of which the final rotation angle is determined. The procedure has been implemented in the clinics at UCLA. The rate of correctly rotated images is 95.4%.

The CLEF 2004 Cross-Language Image Retrieval Track

“…However, previous experiments have been restricted to a small number of categories. For instance, several algorithms have been proposed for orientation detection of chest radiographs, where lateral and frontal orientation are distinguished by means of image content analysis [18,19]. For this two-class experiment, the error rates are below 1 % [20].…”

Automatic categorization of medical images for content-based retrieval and data mining

“…For instance, several algorithms have been proposed for orientation detection of chest radiographs, where lateral and frontal orientation are distinguished by means of digital image processing [22,23]. For this two-class experiment, the error rates are below 1% [24].…”

Determining the View of Chest Radiographs

“…2 To implement a CAD system into the clinical environment, it is important to correctly identify the orientation of image acquisition, ie, posteroanterior (PA) or anteroposterior (AP) versus lateral view, 3,4 as well as top-down, left-right, and mirroring. [5][6][7] These are also essential preprocessing steps for data handling in picture archiving and communication systems (PACS) and content-based image retrieval (CBIR). 8 Pietka and Huang 5 presented an automatic three-step procedure that determined the image orientations of computed radiography chest images and rotated them in steps of 90 degrees into a standard position for viewing by radiologists.…”